Cortex II Data Store for Large and Complex Data.             

Down load overview (300K)

The AURA technology solves the problem of searching large complex data sets using a suite of powerful  tools.  AURA is so scalable that when dealing with massive datasets only an enterprise-level computer system can provide the capability to make full use of its capabilities.

When an IT decision is made to shift to using Enterprise-level computing to support data processing needs, so begins the long, and often painful cycle of mixing and matching the various offerings from the large systems integrators, to get the best performance for the budget.   With ten years of experience in the high-performance computing market, Cybula has put together the Cortex II system. The Cortex II is a matched system constructed  from industry standard servers and peripherals in addition to pattern match nodes it provides the most stable,  powerful, expandable and accessible enterprise server system, ideally suited for the AURA technology, and therefore for parallel pattern matching.

Thanks to the modular design of the AURA components, which stretches from individual processing cores within the Presence II cards, up to grid-enabled server farms, many Cortex II systems may be clustered via optional pattern match control software allowing large distributed data to be searched. Thanks to its expertise in Grid-based computing, a Cortex II server farm can be as distributed as your network infrastructure allows, allowing for complete scalability. 

The power of the Cortex II does not stop with AURA.  individual compute nodes (PRESENCE II) have been designed to support AURA search operations, at its heart is the industry-standard Xylinx Virtex II, a fully reprogrammable FPGA chip.  Supporting the Xylinx is a Texas-Instruments’ TMS320 Digital Signal Processor, also fully programmable using standard tools, and up to four gigabytes of on board memory per card.

Providing unrivalled access to large complex datasets, Cortex II is the new standard in high-performance data-rich computing.  The system is primarily aimed at applications using huge amounts of data that can be noisy and potentially incomplete.  Thanks to its flexibility and scalability, the Cortex II is extremely well-placed to rival similar pattern-matching systems.

 AURA Applications areas include:

• Financial data analysis
• Diagnostic systems
• Medical signal analysis
• Address management and lookup
• Image databases
• Document management
• Access to Pharmaceutical data

Cortex II complements existing database servers by providing fast and effective access to both complex data (i.e. graphs, numbers & text) as well as data that is incomplete or noisy.  Typically this type of data is difficult to index with conventional databases.

Cybulas application adapters build on the base data types to allow searching of specific data. Existing systems include trade mark searching (AURAtm), molecular searching (AURAmol), signal data searching (Signal Data Explorer), text searching (AURAtse), document searching (MinerTaur), face searching (FaceEnforce). Other adaptors may be developed by the user or by Cybula.

Utilising Cybula’s own high performance pattern match engine, AURA, this highly flexible parallel computer applies a unique design with methods that allow access to large, complex and uncertain data.  A system can be configured with any number of these unique nodes depending on the performance required by the user, offering scalability that can meet current and future demands for any organisation.

The AURA technology is aimed at customers wishing to integrate flexible access to complex data within large data-rich applications.  Developers are presented with an easy-to-use API that can be used to integrate with existing applications, and Cortex II is the cost-effective parallel system designed to support these applications.

Cybula provide comprehensive training and documentation as well as integration consultancy to allow users to quickly apply the system to solve their data problems.

Features:

 Cortex II can be used effectively for large and small datasets.

 Cortex II AURA software supports three main types of data:

·        Numbers: i.e. signals, measurements.

·        Text: i.e. documents, lists, web pages.

·        Graphs: i.e. images and knowledge structures.

Cortex II can be used stand alone, on a network or within a distributed computing environment (Grid’s).

 Cortex II is hosted on industry-standard platforms.

 A Cortex II system uses Cybula’s unique high-performance pattern match PCI hardware, PRESENCE 2.  A system may be configured for any number of PRESENCE 2 cards and AURA Pattern Match Engines.

 Each PRESENCE 2 node contains up to 4Gb of on board memory for data structures. 

Example configuration for very high performance computing:

For more information on Cortex II please contact the Cybula Sales team.

home : products : services : news : the company : partners : contact